Guides

GEO vs SEO: what changes for a small agency

Written by

GeoShark

Last updated: 21 September 20264 min readReviewed by GeoShark Team

Share

SEO earns a page a position in a list of links. GEO earns a brand a place inside an answer an AI writes. The fundamentals overlap, and Google says the same SEO practices apply to its AI features. What changes is measurement, because an answer has no rank to report, and where the work happens, because engines cite pages you do not own.

Key takeaways

  • Google states there are no additional requirements to appear in AI Overviews or AI Mode. A page has to be indexed and eligible to show with a snippet, the same as for any result.
  • Ranking well helps less than it did. Ahrefs found 76% of cited pages ranked in the top 10 in July 2025 and 38% in a 2026 update with a different method.
  • There is no rank to report. You measure whether a brand is named and recommended across a set of questions, and you repeat it.
  • A large share of the work sits off your client's site, on the review pages, lists and forums the engines read.
  • You do not need a new service line. You need a new deliverable: a report that answers the question 'does AI recommend us?'.

If you sell SEO, your clients have started asking about ChatGPT. The honest answer to "is this a different job?" is: mostly no, with two exceptions that matter.

What carries over

Google is unusually direct about this. Its documentation on AI features says you can apply the same foundational SEO best practices as for Search overall, and that there are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary. It adds that you don't need new machine readable files, AI text files or markup.

The eligibility rule is the ordinary one. A page has to be indexed and eligible to be shown in Google Search with a snippet. Crawlable, indexable, useful, and written for people. Google's own guidance on helpful content asks for content created primarily for people, and not to manipulate search engine rankings.

So the audit you already run still applies. A page blocked from crawling cannot be cited, and a thin page gives an engine nothing to lift. Keep doing that work.

What changes: measurement

A search result has a position. An AI answer is a paragraph, and a brand is either in it or not.

That breaks the usual report. "We moved from position 6 to position 4" has no equivalent. What you can report is a rate: across these 40 buyer questions, on these four engines, the brand was named in this share of answers and recommended in this share. And because the same question returns different answers on different days, one reading proves little. Why the score moves between runs is worth reading before you put a number in front of a client.

What changes: where the work happens

Ranking is a property of a page you control. Being cited is a property of a page an engine chose, and that page is often not yours.

Ahrefs looked at this in 2025 and found 76% of pages cited in AI Overviews ranked in the top 10. In a 2026 update covering 863,000 keywords and four million AI Overview URLs, the figure was 38%, with 31.2% of citations from pages ranking 11 to 100 and 31.0% from beyond position 100. The method changed between the two studies, so do not quote the drop as a precise number. The direction is the useful part: a top-10 ranking is a weaker signal of being cited than it was.

Take an example from our own runs. In a measured run on Nike (August 2026, ten buyer questions), the brand was named in nine of them, yet its own website was cited twice out of 341 sources the engines read, about 0.6%. The engines were talking about Nike using other people's pages. That is one brand on one day, a snapshot and not a category statistic, but it shows the shape: a strong brand with little ownership of the sources.

The pages that fill that gap are review sites, comparison lists, forums and video. Reading the sources behind an answer tells you which ones, and it is the step SEO reporting has no habit for.

What to sell

You do not need a new service line, and you should be wary of anyone selling one as a mystery. You need a new deliverable.

  1. A baseline: the questions a client's buyers ask, run across the engines, with the brand's mention and recommendation rates beside two or three competitors.
  2. The sources: which pages were cited, and how many belong to the client.
  3. The work: pages to publish on the client's site, and the third-party pages worth pitching.
  4. The re-run, to show what moved.

That is a monthly report with something in it. How to write a GEO report for clients sets out the structure, and How to choose an AI visibility tool for agencies covers what to look for if you would rather not collect the answers by hand.

A caution on the research

The best-known academic result here comes from the GEO paper, which found that adding cited sources, quotations and statistics raised a page's visibility in generated answers, and that keyword stuffing did not help. It was tested on a research benchmark, not on your client's category. Treat it as a reason to write plainly and cite properly, not as a guarantee.

If you are new to the term, what GeoShark and GEO are is the short version. For the measurement side start with how to track your brand in AI search answers.

FAQ

Is GEO the same as SEO?
No, but they overlap heavily. Google says the same foundational SEO practices apply to its AI features and that no special optimization is required. GEO adds a different thing to measure (whether a brand is named inside a generated answer) and a different place to work (the third-party pages an answer is built from).
Do I need llms.txt or special markup to appear in AI Overviews?
Not for Google. Its documentation says you don't need to create new machine readable files, AI text files, or markup to appear in AI Overviews or AI Mode. Other engines publish their own guidance, so check each one rather than assuming.
Will ranking in the top 10 get a client cited in AI Overviews?
It helps, but less than it did. Ahrefs reported that 38% of pages cited in AI Overviews ranked in the top 10 for the query in a 2026 study of 863,000 keywords, against 76% in its July 2025 study. The two studies differ in method, so read the drop as a direction rather than an exact fall.
How do I know if a client's AI visibility work is paying off?
Ask the same set of buyer questions on a schedule and record whether the brand is named, whether it is recommended, and which sources the answers cite. Movement across repeated runs is the evidence. Google Search Console reports AI feature traffic inside the Web search type, so it cannot tell you the AI share on its own.

Sources

  1. AI features and your website — Google Search Central, 2026
  2. Google AI Overview citations from top-ranking pages drop sharply — Search Engine Journal, reporting Ahrefs data, 2026
  3. 76% of AI Overview citations pull from the top 10 pages — Ahrefs, 2025
  4. GEO: Generative Engine Optimization (Aggarwal et al.) — arXiv, 2024
  • GEO
  • SEO
  • agencies

New to this? Start with what GeoShark and GEO are.